Two-Sided Bandits and the Dating Market
Speaker: Sanmay Das , CSAIL, MIT
Date: November 15 2004
Time: 4:10PM to 5:00PM
Location: Patil/Kiva seminar room 32-G449
Contact: Louis-Philippe Morency, 617-253-4278, email@example.com
We define a "two-sided bandit problem" as the decision problem facing agents in repeated matching environments with learning, and examine the dating market, in which men and women repeatedly go out on dates and learn about each other, as an example of such an environment. Three natural matching mechanisms are:
1. Gale-Shapley matching, in which each agent submits a list of preferences over those on the other side of the market in each period and a centralized algorithm computes a stable matching.
2. Simultaneous choice, in which each agent on side 1 submits an offer to one agent on side 2, and the agents on side 2 choose which offer to accept at the end of the period.
3. Sequential choice, in which the agents on side 1 are randomly ordered in each period and then submit "take it or leave it" offers to those on side 2.
We empirically examine the properties of these mechanisms, focusing on the asymptotic stability of the resulting matchings, and consider "optimistic" algorithms that increase the probability of stable outcomes.
Note: Joint work with Emir Kamenica, Harvard
See other events that are part of CSAIL Student Seminar Series Fall 2004
See other events happening in November 2004